Clustering & Cell Type Annotation

Ramon Massoni-Badosa

Note: The key parameter in this clustering analysis is the resolution in FindClusters. This was set after iterative trials of this pipeline combined with our domain knowledge of the clusters obtained.

# Key parameters
num_significant_pcs <- 14
k_resolution <- 0.01

1 Introduction

In the previous notebook, we filtered and normalized the expression matrix. Here, we aim to identify the clusters of cells in the dataset and annotate them to their corresponding cell type. Furthermore, we will try to identify cycling cells.

2 Pre-processing

2.1 Package loading

library(scater)
library(scran)
library(Seurat)
library(ggpubr)
library(tidyverse)
library(kBET)
library(cluster)
library(DoubletFinder)

2.2 Source script with functions

source("bin/utils.R")

2.3 Load data

cll_seu <- readRDS("results/R_objects/cll_seurat_filtered_normalized.rds")
cll_seu$barcode <- colnames(cll_seu)

3 Find Variable Genes

To cluster our cells, we need to overcome 2 challenges:

1. The ‘curse of dimensionality’: as each cell can be conceived as a vector with >10,000 genes, and as two random cells will have the very similar expression for most genes, the distance measured between any given pair of cells will be very low, thus being unreliable for proper comparisons.
2. Computational complexity: as the data is highly dimensional, even the most greedy algorithm will take long to complete.
3. Most genes should not be differentially expressed between cells, so the observed differences in such genes will be due to technical issues or transient biological states, hence confounding the true structure in the dataset.

Thus, we aim to eliminate redundancy and denoise the dataset. To this end, we will find the subset of genes that drive most of the variability in the expression matrix (feature selection):

cll_seu <- FindVariableFeatures(cll_seu)
num_var_genes <- length(VariableFeatures(cll_seu))
num_var_genes

## [1] 2000

var_plot <- VariableFeaturePlot(cll_seu)
var_plot <- LabelPoints(
  plot = var_plot, 
  points =  head(VariableFeatures(cll_seu), 10), 
  repel = TRUE
)
var_plot

As we can see, we reduce the number of dimensions from >10,000 genes to 2000 highly variable genes (HVG). Moreover, we see that among the top HVG there are well-known PBMC markers, such as LYZ, S100A8 (monocytes), GNLY and NKG7 (natural killer and CD8+ T cells).

4 Scale data

An important pre-processing step in any cluster analysis is to scale the data, as otherwise variables with a higher mean will have a higher weight in the distance matrix:

cll_seu <- ScaleData(cll_seu)

5 Linear dimensionality reduction (PCA)

An additional challenge in our cluster analysis is that scRNAs-seq is very noisy (very susceptible to technical artifacts), and very sparse (contains drop-outs). Thus, differences in single genes may not be accurate to identify cell types. To that end, we can perform PCA, as every PC can be conceived as a ‘metagene’ that includes information across a correlated gene set. Furthermore, we will reduce the dimensionality even more:

cll_seu <- RunPCA(cll_seu, features = VariableFeatures(cll_seu))
VizDimLoadings(cll_seu, dims = 1:3, reduction = "pca")

DimPlot(cll_seu, reduction = "pca")

6 Determine statistically significant principal components

To determine the number of significant PCs to use, we will use an elbow plot:

ElbowPlot(cll_seu)

The elbow is in PC 14.

7 Cluster cells

To cluster cells we will used the Seurat’s built-in functions FindNeighbors and FindClusters, which use the graph-based Louvain algorithm.

cll_seu <- FindNeighbors(cll_seu, dims = 1:num_significant_pcs)
cll_seu <- FindClusters(cll_seu, resolution = k_resolution)

## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
## 
## Number of nodes: 37006
## Number of edges: 1122570
## 
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.9968
## Number of communities: 5
## Elapsed time: 9 seconds

8 Non-linear dimensionality reduction

We can visualize the former clusters with a t-Stochastic Neighbor Embedding (tSNE) and uniform manifold approximation and projection (UMAP), which allow to depict more structure in the data than PCA:

cll_seu <- RunTSNE(cll_seu, reduction = "pca", dims = 1:num_significant_pcs, check_duplicates = FALSE)
cll_seu <- RunUMAP(cll_seu, reduction = "pca", dims = 1:num_significant_pcs)
DimPlot(cll_seu, reduction = "tsne")

DimPlot(cll_seu, reduction = "umap")

num_k <- length(levels(cll_seu$seurat_clusters))
num_k

## [1] 5

As we can see, there are 5 major clusters.

9 Find cluster markers

Let us find the markers for each of the clusters. That is, the genes that are exclusively expressed in one cluster:

markers <- FindAllMarkers(
  cll_seu,
  features = VariableFeatures(cll_seu),
  only.pos = TRUE
)
DT::datatable(markers)

# saveRDS(markers, "results/R_objects/markers_cll_clusters.rds")
marker_selection <- unlist(map(0:(num_k-1), ~markers[markers$cluster == ., "gene"][0:5]))
marker_selection

##  [1] "IGLC2"    "IGHM"     "IGLC3"    "HLA-DRB5" "IGHD"     "IGHG1"    "PLCG2"    "IGKC"     "TFEC"     "COBLL1"   "IGHA1"    "APOD"     "IGHA2"    "IGKC"     "NOP53"    "GNLY"     "NKG7"     "CCL5"     "IL32"     "IL7R"     "S100A9"   "LYZ"      "S100A8"   "AIF1"     "FCN1"

DoHeatmap(cll_seu, features = marker_selection) + NoLegend()

As we know from the initial PBMC study, cluster 4 are monocytes (as they express LYZ + S100A8); whilst cluster 3 are T and NK cells (as they express NKG7, GNLY and IL7R). We can project these markers in UMAP space:

FeaturePlot(
  cll_seu, 
  features = c("LYZ", "S100A8", "NKG7", "GNLY", "IL7R", "MS4A1"), 
  reduction = "umap"
)

Clusters 0, 1 and 3 are leukemic cells, as they express immunoglobulins (IGLC2, IGHM, IGLC3, IGHD, IGHG1, IGKC, IGHA1 and IGHA2). This is consistent with the fact that CLL cells have a constitutively active B cell Receptor (BCR) signaling.

FeaturePlot(cll_seu, features = c("IGLC2", "IGKC", "IGHA1"), reduction = "umap")

CLL possesses a vast inter-patient variability, and thus we expect clusters 0, 1 and 2 to be from the three different donors (1220, 1472, 1892):

Idents(cll_seu) <- "donor"
DimPlot(cll_seu, reduction = "umap")

10 Cell cycle scoring

We can detect dividing cells with cell-cycle signatures:

s_genes <- cc.genes$s.genes[cc.genes$s.genes %in% rownames(cll_seu@assays$RNA@scale.data)]
g2m_genes <- cc.genes$g2m.genes[cc.genes$g2m.genes %in% rownames(cll_seu@assays$RNA@scale.data)]
cll_seu <- CellCycleScoring(object = cll_seu, s.features = s_genes, g2m.features = g2m_genes)
FeaturePlot(object = cll_seu, features = c("S.Score", "G2M.Score"))

Although one would expect that leukemic cells are cycling, we have to keep in mind that this is a chronic disease were cells accumulate during decades. Thus, at a single time-point it is hard to detect differencesin cell proliferation. Moreover, CLL usually proliferate in lymph nodes, not in the blood

11 Assigning cell type identity to clusters

Based on the previous findings, we can annotate each cluster to known cell types:

Cluster ID	Markers	Cell Type
0	IGLC2	CLL 1892
1	IGKC	CLL 1472
2	IGHA1	CLL 1220
3	GNLY, NKG7, IL7R	T and NK
4	LYZ, S100A	Monocytes

Idents(cll_seu) <- "seurat_clusters"
new_cluster_ids <- c("CLL 1892", "CLL 1472", "CLL 1220", "T and NK", "Monocyte")
names(new_cluster_ids) <- levels(cll_seu)
cll_seu <- RenameIdents(cll_seu, new_cluster_ids)
cll_seu$cell_type <- Idents(cll_seu)
DimPlot(cll_seu, reduction = "umap", label = TRUE, pt.size = 0.5) + NoLegend()

12 Doublet detection

Although conceptually this step would belong to the quality control and normalization notebook, we will proceed to detect the doublets present in our dataset. We carry it out here because we will use DoubletFinder, a tool that works with a fully processed Seurat object.

DoubetFinder simulates doublets by averaging the UMI of two real cells. It subsequently project the simulations and real cells in PCA space and computes a nearest neighbors graph. Doublets are identified as those cells that have a proportion of artificial neighbors (pAN) greater than what you would expect by chance.

Note that different donors and temperatures were sequenced in different 10X lanes. Hence, as it is impossible to find doublets from two different sequencing libraries, we will split the object by library:

# Create list of seurat objects (split by library)
cll_seu$library <- str_c(cll_seu$donor, cll_seu$temperature, sep = "_")
cll_seu_l <- SplitObject(cll_seu, split.by = "library")

# Preprocess independently
cll_seu_l <- purrr:::map(cll_seu_l, function(seurat) {
  seurat %>% 
    NormalizeData() %>% 
    FindVariableFeatures() %>%
    ScaleData() %>% 
    RunPCA() %>% 
    RunTSNE(dims = 1:15) %>% 
    RunUMAP(dims = 1:15)
})
umaps_l <- purrr::map(cll_seu_l, DimPlot, reduction = "umap")
ggarrange(plotlist = umaps_l, nrow = 3, ncol = 2)

The most important parameter to tune when using DoubletFinder is pK, which is the PC neighborhood size used to compute the proportion of artificial neighbors (pANN), as explained in the github page of the package. We can optimize this parameter as follows:

# pK identification
pk_dbl <- purrr::map_dbl(cll_seu_l, function(seurat) {
  sweep_res_list <- paramSweep_v3(seurat, PCs = 1:10, sct = FALSE)
  sweep_stats <- summarizeSweep(sweep_res_list, GT = FALSE)
  bcmvn <- find.pK(sweep_stats)
  pK <- as.numeric(as.character((bcmvn[which.max(bcmvn$BCmetric), "pK"])))
  pK
})

## [1] "Creating artificial doublets for pN = 5%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 10%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 15%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 20%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 25%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 30%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."

## NULL
## [1] "Creating artificial doublets for pN = 5%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 10%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 15%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 20%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 25%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 30%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."

## NULL
## [1] "Creating artificial doublets for pN = 5%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 10%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 15%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 20%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 25%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 30%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."

## NULL
## [1] "Creating artificial doublets for pN = 5%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 10%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 15%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 20%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 25%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 30%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."

## NULL
## [1] "Creating artificial doublets for pN = 5%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 10%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 15%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 20%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 25%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."
## [1] "Creating artificial doublets for pN = 30%"
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Defining neighborhoods..."
## [1] "Computing pANN across all pK..."
## [1] "pK = 0.005..."
## [1] "pK = 0.01..."
## [1] "pK = 0.02..."
## [1] "pK = 0.03..."
## [1] "pK = 0.04..."
## [1] "pK = 0.05..."
## [1] "pK = 0.06..."
## [1] "pK = 0.07..."
## [1] "pK = 0.08..."
## [1] "pK = 0.09..."
## [1] "pK = 0.1..."
## [1] "pK = 0.11..."
## [1] "pK = 0.12..."
## [1] "pK = 0.13..."
## [1] "pK = 0.14..."
## [1] "pK = 0.15..."
## [1] "pK = 0.16..."
## [1] "pK = 0.17..."
## [1] "pK = 0.18..."
## [1] "pK = 0.19..."
## [1] "pK = 0.2..."
## [1] "pK = 0.21..."
## [1] "pK = 0.22..."
## [1] "pK = 0.23..."
## [1] "pK = 0.24..."
## [1] "pK = 0.25..."
## [1] "pK = 0.26..."
## [1] "pK = 0.27..."
## [1] "pK = 0.28..."
## [1] "pK = 0.29..."
## [1] "pK = 0.3..."

## NULL

Now we are equipped to run DoubletFinder:

cll_seu_l <- map2(cll_seu_l, pk_dbl, function(seurat, pk) {
  nExp_poi <- round(0.1 * ncol(seurat))
  seurat <- doubletFinder_v3(
    seu = seurat, 
    PCs = 1:10, 
    pN = 0.25, 
    pK = pk, 
    nExp = nExp_poi, 
    reuse.pANN = FALSE, 
    sct = FALSE
  )
})

## [1] "Creating 2205 artificial doublets..."
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Computing pANN..."
## [1] "Classifying doublets.."
## [1] "Creating 1319 artificial doublets..."
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Computing pANN..."
## [1] "Classifying doublets.."
## [1] "Creating 2971 artificial doublets..."
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Computing pANN..."
## [1] "Classifying doublets.."
## [1] "Creating 2704 artificial doublets..."
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Computing pANN..."
## [1] "Classifying doublets.."
## [1] "Creating 3136 artificial doublets..."
## [1] "Creating Seurat object..."
## [1] "Normalizing Seurat object..."
## [1] "Finding variable genes..."
## [1] "Scaling data..."
## [1] "Running PCA..."
## [1] "Calculating PC distance matrix..."
## [1] "Computing pANN..."
## [1] "Classifying doublets.."

We can add the annotations back to the original Seurat object:

doublet_annot <- purrr::map(cll_seu_l, function(seurat) {
 selected_column <- str_subset(
   string = colnames(seurat@meta.data), 
   pattern = "^DF.classifications"
 )
 doub_find <- seurat@meta.data[, selected_column, drop = FALSE] 
 colnames(doub_find) <- "doublet_annot"
 doub_find <- rownames_to_column(doub_find, var = "barcode")
 doub_find
})
doublet_annot <- bind_rows(doublet_annot, .id = "library")
rownames(doublet_annot) <- doublet_annot$barcode
doublet_annot <- doublet_annot[colnames(cll_seu), ]
cll_seu$doublet_annot <- doublet_annot$doublet_annot

# Visualize
Idents(cll_seu) <- "doublet_annot"
DimPlot(cll_seu, reduction = "umap")

As a validation, we can load the demultiplexed and unfiltered seurat object, remove all cells not present in the current object, add the annotation of DoubletFinder, process the object and assess if the cell hashing-derived and DF-derived doublets cluster together:

cll_ori <- readRDS("results/R_objects/cll_Seurat_demultiplexed.rds")
cll_ori$barcode <- colnames(cll_ori)
cll_ori <- subset(cll_ori, hash.ID == "Doublet" | barcode %in% colnames(cll_seu))
cll_ori$hash.ID[cll_ori$hash.ID == "Doublet"] <- "Doublet_hashing"
doublet_finder_barcodes <- colnames(cll_seu)[cll_seu$doublet_annot == "Doublet"]
cll_ori$hash.ID[doublet_finder_barcodes] <- "Doublet_Finder"
singlet_barcodes <- colnames(cll_seu)[cll_seu$doublet_annot == "Singlet"]
cll_ori$hash.ID[singlet_barcodes] <- "Singlet"

# Pre-process
cll_ori <- cll_ori %>% 
  NormalizeData() %>% 
  FindVariableFeatures() %>% 
  ScaleData() %>% 
  RunPCA() %>% 
  RunUMAP(dims = 1:10)
Idents(cll_ori) <- "hash.ID"
DimPlot(cll_ori, reduction = "umap")

As we can see, DoubletFinder-predicted and hashing-predicted doublets cluster together, which validates the prediction of DoubletFinder. Hence, we can clean our dataset even more and filter them out:

cll_seu <- subset(cll_seu, subset = doublet_annot == "Singlet")
cll_seu

## An object of class Seurat 
## 15710 features across 33305 samples within 1 assay 
## Active assay: RNA (15710 features)
##  3 dimensional reductions calculated: pca, tsne, umap

Idea: given that we have ground-truth about between-timepoint doublets, we could find, for each cell, the proportion of neareast doublets, correlate it with highly-expressing genes and discard the outliers.

Now that, we performed the final filtering step in our QC, let us quantify the number of high-quality cells stratified by time, temperature and condition.

qc_df2 <- cll_seu@meta.data
qc_df2$temperature <- case_when(
  qc_df2$temperature == "4C" ~ "4ºC",
  qc_df2$temperature == "RT" & qc_df2$time == "0h" ~ "0h",
  qc_df2$temperature == "RT" & qc_df2$time != "0h" ~ "21ºC"
)
qc_df2 <- qc_df2 %>%
  mutate(time = factor(time, levels = c("0h", "2h", "4h", "6h", "8h", "24h")),
         temperature = factor(temperature, levels = c("0h", "21ºC", "4ºC")),
         donor = factor(donor, levels = c("1220", "1472", "1892"))) %>%
  group_by(donor, temperature, time) %>% 
  summarise(num_cells_after_qc = n())
col_sel <- c("time", "temperature", "donor", "num_cells_after_qc")
qc_df2 <- qc_df2[, col_sel]
qc_df <- readRDS(file = "results/R_objects/qc_summary_table_cll.rds")
qc_df <- add_column(
  qc_df,
  num_cells_after_qc = qc_df2$num_cells_after_qc,
  .after = "num_cells_before_qc"
)
qc_df <- add_column(
  qc_df,
  experiment = rep("CLL", nrow(qc_df)),
  .after = 0
)

# Merge it with PBMC
qc_pbmc <- readRDS("../1-PBMC/results/R_objects/qc_summary_table_pbmc.rds")
qc_pbmc <- add_column(
  qc_pbmc,
  experiment = rep("PBMC", nrow(qc_pbmc)),
  .after = 0
)
qc_df_all <- rbind(qc_pbmc, qc_df)
DT::datatable(qc_df_all)

# saveRDS(qc_df_all, file = "results/R_objects/qc_summary_table_all.rds")

Let us do the same but for each cell type:

qc_cell_type_cll <- cll_seu@meta.data
qc_cell_type_cll$temperature <- case_when(
  qc_cell_type_cll$temperature == "4C" ~ "4ºC",
  qc_cell_type_cll$temperature == "RT" & qc_cell_type_cll$time == "0h" ~ "0h",
  qc_cell_type_cll$temperature == "RT" & qc_cell_type_cll$time != "0h" ~ "21ºC"
)
qc_cell_type_cll <- qc_cell_type_cll %>%
  mutate(temperature = factor(temperature, levels = c("0h", "21ºC", "4ºC")),
         donor = factor(donor, levels = c("1220", "1472", "1892"))) %>%
  group_by(time, donor, temperature, cell_type) %>%
  summarise(num_cells = n()) %>%
  ungroup()
colnames(qc_cell_type_cll)[4] <- "annotation"
qc_cell_type_cll <- add_column(
  qc_cell_type_cll,
  experiment = rep("CLL", nrow(qc_cell_type_cll)),
  .after = 0
)
qc_cell_type_pbmc <- readRDS("../1-PBMC/results/R_objects/qc_summary_table_cell_type_pbmc.rds")
qc_cell_type_df_all <- rbind(qc_cell_type_pbmc, qc_cell_type_cll)
DT::datatable(qc_cell_type_df_all)

# Save as excel file
to_excel <- list(sampling_time = qc_df_all, cell_type = qc_cell_type_df_all)
# openxlsx::write.xlsx(to_excel, file = "results/tables/quality_control_metrics.xlsx")

13 Save Seurat object

# saveRDS(cll_seu, file = "results/R_objects/cll_seurat_annotated.rds")

14 Session Info

sessionInfo()

## R version 3.6.1 (2019-07-05)
## Platform: x86_64-apple-darwin15.6.0 (64-bit)
## Running under: macOS High Sierra 10.13.6
## 
## Matrix products: default
## BLAS:   /Library/Frameworks/R.framework/Versions/3.6/Resources/lib/libRblas.0.dylib
## LAPACK: /Library/Frameworks/R.framework/Versions/3.6/Resources/lib/libRlapack.dylib
## 
## locale:
## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
## 
## attached base packages:
##  [1] grid      parallel  stats4    stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
##  [1] modes_0.7.0                 ROCR_1.0-7                  gplots_3.0.3                KernSmooth_2.23-16          fields_10.3                 maps_3.3.0                  spam_2.5-1                  dotCall64_1.0-0             DoubletFinder_2.0.2         cluster_2.1.0               kBET_0.99.6                 forcats_0.5.0               stringr_1.4.0               dplyr_0.8.4                 purrr_0.3.3                 readr_1.3.1                 tidyr_1.0.2                 tibble_2.1.3                tidyverse_1.3.0             ggpubr_0.2.5                magrittr_1.5                Seurat_3.1.4                scran_1.14.6                scater_1.14.6               ggplot2_3.3.0               SingleCellExperiment_1.8.0  SummarizedExperiment_1.16.1 DelayedArray_0.12.2         BiocParallel_1.20.1         matrixStats_0.55.0          Biobase_2.46.0              GenomicRanges_1.38.0        GenomeInfoDb_1.22.0         IRanges_2.20.2              S4Vectors_0.24.3            BiocGenerics_0.32.0         BiocStyle_2.14.4           
## 
## loaded via a namespace (and not attached):
##   [1] reticulate_1.14          tidyselect_1.0.0         htmlwidgets_1.5.1        Rtsne_0.15               munsell_0.5.0            codetools_0.2-16         mutoss_0.1-12            ica_1.0-2                DT_0.12                  statmod_1.4.34           future_1.16.0            withr_2.1.2              colorspace_1.4-1         knitr_1.28               rstudioapi_0.11          ggsignif_0.6.0           gbRd_0.4-11              listenv_0.8.0            Rdpack_0.11-1            labeling_0.3             GenomeInfoDbData_1.2.2   mnormt_1.5-6             farver_2.0.3             vctrs_0.2.3              generics_0.0.2           TH.data_1.0-10           xfun_0.12                R6_2.4.1                 ggbeeswarm_0.6.0         rsvd_1.0.3               locfit_1.5-9.1           bitops_1.0-6             assertthat_0.2.1         promises_1.1.0           scales_1.1.0             multcomp_1.4-12          beeswarm_0.2.3           gtable_0.3.0             npsurv_0.4-0             globals_0.12.5           sandwich_2.5-1           rlang_0.4.5              splines_3.6.1            lazyeval_0.2.2           broom_0.5.5              BiocManager_1.30.10      yaml_2.2.1              
##  [48] reshape2_1.4.3           modelr_0.1.6             crosstalk_1.0.0          backports_1.1.5          httpuv_1.5.2             tools_3.6.1              bookdown_0.18            RColorBrewer_1.1-2       ggridges_0.5.2           TFisher_0.2.0            Rcpp_1.0.3               plyr_1.8.6               zlibbioc_1.32.0          RCurl_1.98-1.1           pbapply_1.4-2            viridis_0.5.1            cowplot_1.0.0            zoo_1.8-7                haven_2.2.0              ggrepel_0.8.1            fs_1.3.2                 RSpectra_0.16-0          magick_2.3               data.table_1.12.8        lmtest_0.9-37            reprex_0.3.0             RANN_2.6.1               mvtnorm_1.1-0            fitdistrplus_1.0-14      xtable_1.8-4             mime_0.9                 hms_0.5.3                patchwork_1.0.0          lsei_1.2-0               evaluate_0.14            readxl_1.3.1             gridExtra_2.3            compiler_3.6.1           crayon_1.3.4             htmltools_0.4.0          later_1.0.0              RcppParallel_4.4.4       lubridate_1.7.4          DBI_1.1.0                dbplyr_1.4.2             MASS_7.3-51.5            Matrix_1.2-18           
##  [95] cli_2.0.2                gdata_2.18.0             metap_1.3                igraph_1.2.4.2           pkgconfig_2.0.3          sn_1.5-5                 numDeriv_2016.8-1.1      plotly_4.9.2             xml2_1.2.2               vipor_0.4.5              dqrng_0.2.1              multtest_2.42.0          XVector_0.26.0           bibtex_0.4.2.2           rvest_0.3.5              digest_0.6.25            sctransform_0.2.1        RcppAnnoy_0.0.15         tsne_0.1-3               rmarkdown_2.1            cellranger_1.1.0         leiden_0.3.3             uwot_0.1.5               edgeR_3.28.1             DelayedMatrixStats_1.8.0 shiny_1.4.0              gtools_3.8.1             lifecycle_0.1.0          nlme_3.1-145             jsonlite_1.6.1           BiocNeighbors_1.4.2      viridisLite_0.3.0        limma_3.42.2             fansi_0.4.1              pillar_1.4.3             lattice_0.20-40          fastmap_1.0.1            httr_1.4.1               plotrix_3.7-7            survival_3.1-8           glue_1.3.1               FNN_1.1.3                png_0.1-7                stringi_1.4.6            BiocSingular_1.2.2       caTools_1.18.0           irlba_2.3.3             
## [142] future.apply_1.4.0       ape_5.3